2.4.16Thermodynamics & Statistical Mechanics (Advanced)

Bose-Einstein statistics — bosons

1,706 words8 min readdifficulty · medium5 backlinks

WHAT are we counting?


HOW we derive the distribution (from scratch)

We want the average number of bosons nˉi\bar n_i in a single-particle energy level εi\varepsilon_i, for a gas in contact with a heat-and-particle reservoir at temperature TT and chemical potential μ\mu.

Step 1 — Use the grand canonical ensemble. Why? Particle number is not fixed (e.g. photons appear/disappear), so we let energy and particles flow. Each level is treated as an independent system that can hold ni=0,1,2,n_i = 0,1,2,\dots particles.

The grand partition function for one level: Zi=ni=0eβ(εiμ)ni,β=1kBT.\mathcal{Z}_i = \sum_{n_i=0}^{\infty} e^{-\beta(\varepsilon_i - \mu)\,n_i}, \qquad \beta = \frac{1}{k_BT}. Why this form? Each particle in the level costs energy εi\varepsilon_i and "removes" μ\mu from the reservoir; nin_i particles cost (εiμ)ni(\varepsilon_i-\mu)n_i. The Boltzmann weight is eβEe^{-\beta E}.

Step 2 — Sum the geometric series. Let x=eβ(εiμ)x = e^{-\beta(\varepsilon_i-\mu)}. Then Zi=n=0xn=11x,(converges only if x<1).\mathcal{Z}_i = \sum_{n=0}^\infty x^n = \frac{1}{1-x}, \qquad \text{(converges only if } x<1). Why this step? It is a geometric series; this is the algebraic place where the boson "unlimited occupancy" enters — the sum runs to \infty. Convergence demands x<1x<1, i.e. εiμ>0\varepsilon_i-\mu>0, i.e. μ<εi\mu<\varepsilon_i for every level. (For fermions the sum stops at n=1n=1, so no such restriction arises.)

Step 3 — Get the mean occupancy. The average nˉi=nnP(n)\bar n_i = \sum_n n\, P(n) where P(n)=xn/ZiP(n)=x^n/\mathcal{Z}_i. A neat trick: nˉi=xxlnZi=xx[ln(1x)]=x1x.\bar n_i = x\frac{\partial}{\partial x}\ln \mathcal{Z}_i = x\frac{\partial}{\partial x}\big[-\ln(1-x)\big] = \frac{x}{1-x}. Why this step? xxlnZx\,\partial_x \ln\mathcal Z is the standard generating-function trick that pulls out n\langle n\rangle from a partition sum.

Step 4 — Substitute back x=eβ(εiμ)x=e^{-\beta(\varepsilon_i-\mu)}.   nˉi=1eβ(εiμ)1  \boxed{\;\bar n_i = \frac{1}{e^{\beta(\varepsilon_i-\mu)} - 1}\;}

Figure — Bose-Einstein statistics — bosons

Reading the formula


Steel-manned mistakes


Recall Feynman: explain it to a 12-year-old

Imagine a stadium where every fan is allowed to crowd into the same seat — there's no "one person per seat" rule. When it's cold (low energy), fans pile up into the cheapest front-row seat. The formula 1/(e(εμ)/kBT1)1/(e^{(\varepsilon-\mu)/k_BT}-1) just tells you how many fans you'll typically find in a seat of a given price ε\varepsilon. Cheap seats → big crowd; expensive seats → almost empty. There's a catch: the "discount" μ\mu can get close to the cheapest seat's price but never equal it — if it did, the maths says infinitely many fans rush in, which is the real-life "Bose-Einstein condensation." Photons of light, sound vibrations, and certain super-cold atoms are these "pile-up" particles.


Flashcards

What spin do bosons have?
Integer spin (0,1,2,0,1,2,\dots).
What symmetry does a boson's total wavefunction obey?
Symmetric under exchange: ψ(1,2)=+ψ(2,1)\psi(1,2)=+\psi(2,1).
State the Bose-Einstein distribution.
nˉ=1e(εμ)/kBT1\bar n=\dfrac{1}{e^{(\varepsilon-\mu)/k_BT}-1}.
Why is occupancy unlimited for bosons?
Symmetric wavefunction doesn't vanish when states coincide → geometric sum runs n=0n=0\to\infty.
Which sign in the denominator distinguishes BE from FD?
BE has 1-1; Fermi-Dirac has +1+1.
What constraint must μ\mu satisfy for bosons, and why?
μ<εmin\mu<\varepsilon_{\min} strictly — needed for the geometric sum 1/(1x)1/(1-x) to converge. At μ=εmin\mu=\varepsilon_{\min} the ground-state partition function diverges ⇒ BEC.
What is μ\mu for photons and why?
μ=0\mu=0, because photon number is not conserved.
Limit recovering Maxwell-Boltzmann?
When εμkBT\varepsilon-\mu\gg k_BT, then nˉe(εμ)/kBT\bar n\approx e^{-(\varepsilon-\mu)/k_BT}.
From Zi=xn\mathcal Z_i=\sum x^n, how do you extract nˉi\bar n_i?
nˉi=xxlnZi=x/(1x)\bar n_i=x\,\partial_x\ln\mathcal Z_i = x/(1-x).
Plugging μ=0, βε=1\mu=0,\ \beta\varepsilon=1 gives what nˉ\bar n?
1/(e1)0.581/(e-1)\approx0.58.

Connections

Concept Map

has

allows

counted by

used because

sum to infinity in

defines

geometric series

requires convergence

apply x d/dx ln Z

substitute x

formula

Boson integer spin

Symmetric wavefunction

Unlimited occupancy n=0..inf

Bose-Einstein statistics

Grand canonical ensemble

Particle number not fixed

Grand partition function Zi

Zi = 1 over 1-x

mu less than epsilon

mean occupancy n = x over 1-x

n = 1 over exp of beta times epsilon-mu minus 1

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Bosons woh particles hote hain jinka spin integer (0,1,2,...) hota hai aur unka wavefunction exchange ke under symmetric hota hai. Iska seedha matlab: jitne marzi bosons ek hi quantum state mein baith sakte hain — koi Pauli exclusion nahi. Yahi "social" nature unko fermions se alag banata hai.

Distribution nikalne ke liye hum grand canonical ensemble use karte hain, kyunki photons jaise bosons ki number conserve nahi hoti. Ek level ke liye partition function n=0xn=1/(1x)\sum_{n=0}^\infty x^n = 1/(1-x) ban jaata hai, jahan x=e(εμ)/kBTx=e^{-(\varepsilon-\mu)/k_BT}. Ye infinite sum sirf tab converge karta hai jab x<1x<1, yani μ<ε\mu<\varepsilon har level ke liye. Average nikalne par milta hai nˉ=1/(e(εμ)/kBT1)\bar n = 1/(e^{(\varepsilon-\mu)/k_BT}-1) — denominator mein minus 1, yahi BE ka signature hai.

Iska physics: jab ε\varepsilon bahut zyada ho to 1-1 ignore ho jaata hai aur classical Maxwell-Boltzmann mil jaata hai. Jab εμ\varepsilon \to \mu ho to denominator zero ki taraf jaata hai aur nˉ\bar n phat jaata hai — yani bosons low-energy state mein crowd kar dete hain. Photons ke liye μ=0\mu=0, aur isi se Planck ka blackbody law banta hai. Important: μ\mu kabhi bhi lowest energy ke barabar bhi nahi ho sakta — strictly μ<εmin\mu<\varepsilon_{\min}. Agar μ=εmin\mu=\varepsilon_{\min} ho jaye to ground-state ka partition function 1/(1x)1/(1-x) hi diverge ho jaata hai, aur yahi BEC ka onset hai.

Go deeper — visual, from zero

Test yourself — Thermodynamics & Statistical Mechanics (Advanced)

Connections